Abstract: The current workflow models are hard to describe uncertain factors in applications before run-time.For this problem，a flexible workflow modeling technique based on so-called flexible block is proposed.The flexible block is running in Ad hoc manner firstly，then it’s inside process structure is gotten through process mining after the running history data being accumulated to a degree.And then，the application scenarios are matched，and the process model of flexible block is educed through interactive case-based reasoning.Practices indicate the technique proposed by this paper can reuse the modeling knowledge very well.Modeling efficiency of flexible block is improved along with the executing times of it.The workflow models containning flexible blocks are with high flexibility.

Key words: workflow model, flexible block, interactive case-based reasoning, flexibility

摘要： 针对工作流模型难以描述创建时的不确定因素的问题，研究了一种基于“柔性块”的柔性工作流建模技术。运行时，柔性块先以Ad hoc方式执行，运行历史数据积累到一定程度后，通过过程挖掘得到柔性块内部模型，并在以后执行时匹配应用场景，采用交互式基于案例的推理方法获得柔性块模型。实际应用表明该方法能复用建模知识，柔性块建模效率随着其执行次数增多而提高，工作流模型具有较高的柔性。

关键词: 工作流模型, 柔性块, 交互式基于案例的推理, 柔性